TTU138: Matt Dorsten, Portfolio Manager, Quantitative Strategies at PIMCO

[00:00:06] Matt

The most important thing I would mention, aside from the defensive profile of trend following, is just the universe. It’s one thing I think people should really focus on; making sure when they’re investing in trend following that they’re getting material exposure to a very broad universe because, as I mentioned, trends being relatively rare, your best chance of success is having exposure to these strange corners of the market. I don’t mean just having them in the portfolio in some token amount, but having material exposure to things like European power or Hungarian interest rates when trends pop up. So, I think that’s a very important thing to keep in mind.

[00:00:44] Intro

Imagine spending an hour with the world’s greatest traders. Imagine learning from their experiences, their successes and their failures. Imagine no more. Welcome to Top Traders Unplugged, the place where you can learn from the best hedge fund managers in the world so you can take your manager due diligence or investment career to the next level.

Before we begin today’s conversation, remember to keep two things in mind. All the discussion we will have about investment performance is about the past, and past performance does not guarantee or even infer anything about future performance. Also understand that there’s a significant risk of financial loss with all investment strategies, and you need to request and understand the specific risks from the investment manager about their product before you make investment decisions.

Here’s your host, veteran hedge fund manager, Niels Kaastrup-Larsen.

[00:01:34] Niels

Hey everyone, and welcome to another edition of Top Traders Unplugged, where today, Alan Dunne and I joined by Matt Dorsten, Portfolio Manager within Quantitative Strategies at PIMCO as part of our mini-series focusing on the one investment strategy that beat everything else in 2022, namely, trend following and managed futures more broadly.

So first off, Matt, it is great to have you on the podcast. Thanks so much for joining us today. We really have been looking forward to our conversation. I hope you’re doing well on the West Coast.

[00:02:07] Matt

Yeah, thanks for having me. I’m very excited. We’re certainly doing very well on the West Coast. We’ve had a lot of unusual snow, but other than that, it’s smooth sailing.

[00:02:15] Niels

Yeah, that’s what I was watching on the television. I’m thinking, well, you complain often about not having enough moisture in your weather, and I’m thinking this year, you may not complain about that.

[00:02:25] Matt

Now we’ve got it. Yeah, we’ve got a good snow pack.

[00:02:28] Niels

Exactly.

Now, before we dive into all the different topics we’re going to talk about with you, I would like just to set the stage a little bit for our conversation, and also so that the audience can know a little bit about PIMCO and especially PIMCO’s involvement in the trend following space. So perhaps you could give us a little bit of a few highlights in terms of the strategies you run within your quantitative group and where the business stands as we’ve entered into 2023.

[00:03:00] Matt

Yeah, so PIMCO, obviously, not something that most people think about when they think about quant investing but we like to think that we’ve been a fairly systematic investor for decades actually. There are papers that Bill Gross wrote two decades ago talking about harvesting structural risk premia, and the difference between treasury bond futures and the actual cash bonds, or harvesting option premia by selling interest rate options, stuff like that. So we take a very systematic approach to bond investing.

Our first foray into your direct systematic investing actually came about 15 years ago with the advent of tail risk hedging, and then shortly after that, we got into trend following and we’ve been focused on that ever since. We’ve broadened out our offerings to more absolute return mandates focused on a wide set of systematic strategies. So we run about $60 billion in assets right now in what we would call quant investing. A large chunk of that is tail risk hedging but the second biggest thing is trend following and that’s what I focus my time on.

[00:04:12] Niels

Unless it’s confidential, of course, what does that leave in trend in terms of AUM that you run in these strategies you would say roughly?

[00:04:21] Matt

It’s on the order of $5 billion.

[00:04:24] Niels

Okay. Very good. Excellent. It sounds like you joined the firm more or less at the time when all this systematic expansion was taking place. Is that right?

[00:04:34] Matt

Yeah, that’s about right.

[00:04:35] Niels

Not to date you, Matt. Not to date you of course.

[00:04:38] Matt

Yeah. When I joined the quant desk, as we call it, it was really just starting. We had tail risk hedging, but we were just starting in trend following.

[00:04:47] Niels

Fantastic. Good stuff.

Now, as you may know, from listening to other episodes, Alan and I have a long list of different topics. We don’t always get to talk about all of them, but we do bring them up and kick them off one by one. So as the audience now will be completely familiar with, I will kick it over to Alan to start off our conversation on these topics today. So over to you, Alan.

[00:05:13] Alan

Thanks, Niels. Matt, good to speak to you. It’s interesting to hear about that history of, as you say, the evolution of quant trading in PIMCO from systematic trading in bonds to tail risk to trend. The fact that you approached trend starting from tail risk, does that say something about how you view trend following and does that say something about the overall philosophy that you bring to trend following?

[00:05:44] Matt

Yeah, that’s very intentional, actually. We started with tail risk at hedging, as I said, with a strong focus on providing clients with defensive protection. In the process of that evolution, we realized, “Hey, there’s this other strategy out here which certainly doesn’t provide guaranteed protection in tail events, like tail risk hedging does, but it seems to have a very consistent defensive profile across time in many different crises.” It’s not guaranteed, it doesn’t happen every time, but that seems to be pretty consistent.

In contrast to tail risk hedging, it doesn’t cost money. In fact, it pays you money to be in it. So we viewed that as a very valuable thing, a very unique strategy for investors and there are very few other things like that out there. So yeah, that’s always informed our view and we’ve tried to enhance that, where we can, in our approach to trend following to really bring out that characteristic because we view that as the most important thing for end investors.

[00:06:43] Alan

In constructing the portfolio, is that very much at the forefront of your minds in terms of building a trend portfolio that is very heavy on that characteristic or are you still very much first and foremost thinking about absolute returns with the additional benefit of that defensive property?

[00:07:02] Matt

I would say first and foremost is the defensive property, and then absolute returns come after that. So one thing that we’re very careful to do is, when we’re investigating new strategies, we may have an idea for a strategy that makes money and looks very good. The first thing that we look at, before adding it to the portfolio is, what does it do to the defensive characteristics?

If it’s not defensive, if it’s reducing that behavior, we have a very high bar to adding things. So there are many things that we think are trend following in nature, thought had a lot of promise, and we’ve not added them to the portfolio because they made the defensive characteristics worse.

[00:07:38] Alan

Okay. That’s very interesting. Most if not all of the managers we’ve spoken to would start off with the premise of philosophically believing that markets trend, and trend following makes sense from that perspective, and over time, have figured out that it has this defensive characteristic. But you very much started off trying to build a portfolio that would give a defensive property and it just so happens that a trend following approach lends itself very well to that.

[00:08:03] Matt

Well, like I said, that characteristic, something that doesn’t cost money or pays you money, in fact, and then is also negatively correlated with equities or defensive in nature, that’s extremely rare in the asset allocation context. It’s extremely valuable if you think it’s going to continue. There are plenty of other strategies out there that add returns and clients do that separately, but the thing that’s really unique and valuable to them is this defensive behavior.

[00:08:31] Alan

From your perspective, why do you think that persists over time? You mentioned that defensive property and how it’s not guaranteed, but it has been fairly consistent through multiple types of crises. Why do you think it has that consistency?

[00:08:49] Matt

Well, I think, there are two ways of looking at it. One more of a quantitative way and one more like a qualitative way. In the qualitative sense, trend following is one of the few strategies that can really get short assets when that’s the trend. Many people can get neutral, but they’re not going short a lot of assets, and they’re certainly very reluctant to do so even if they can.

So when a crisis event is happening and you’ve got a prolonged downturn in assets, the trend follower will jump on that whereas other people at most will probably be flat. On the more mechanical, quantitative side, trend following, I think, no matter how you construct the portfolio, is kind of like replicating a portfolio of long options in different assets. Typically, as the trend goes in your favor, you’re adding to the position, which if you’re replicating an option (a long option position), the delta increases, you need to buy more. So it’s kind of mechanically like a long option position.

So I think it’s very logical that you would expect this convexity type behavior in crisis events. The thing that is surprising is that it doesn’t cost you money to do so. In fact, it pays you money to do so. So that’s the surprising thing that comes back to behavioral anomalies that many investors suffer from.

[00:10:05] Niels

Yeah, I want to stay a little bit with this. It’s very, very interesting and as Alan pointed out, it’s a little bit of a different approach to what we’ve heard from other managers. So my first question would be actually, when you look at your peer group, do you notice that your strategy is behaving differently now that you are coming at it from a slightly different angle maybe even though it is trend following and we’re all trend followers?

But just from a structural point of view, do you notice a difference or, in the end, even though the models might be designed with a slightly different objective, it actually looks very similar to what other trend followers are doing?

[00:10:50] Matt

Well, I think the broad performance behavior is similar. But one thing that we do notice as very quantifiable is the correlation with equities. We’ve realized, over the lifetime of our fund, the most negative correlation with equities of I think all or most of our competitors. So that shows very clearly as confirmation that our approach is working or at least doing what we’re trying to do.

[00:11:13] Niels

Yeah, that’s quite interesting. One of the topics that I like to ask about is this philosophical thinking about do you do pure trend or do you start to diversify your trend following? You’ve clearly come out and said, well, actually, we’re kind of pure camp. That’s great to hear because there aren’t that many of us left in the pure camp. But it brings up this paper that Cliff Asness and Yao from AQR wrote last year that we always like to just hear people’s thoughts about.

One of the things that they mentioned was, are we becoming too concerned about the Sharpe? So even as a pure trend follower, even as someone who takes the stance from, yeah, we definitely need to have some more intention behind providing returns during a “crisis”, how do you look at these risk adjusted metrics? How much weight do you put in them when you design your overall strategy? Are they important to you? How do you look at that?

[00:12:26] Matt

Performance is important. We definitely want to be able to produce something that’s defensive as well as high returning. But like I said, the primary driver is defensiveness.

[00:12:38] Niels

Sorry to interrupt you, Matt. I’m thinking here about the risk adjusted performance, so to speak. To use a really bad example, it’s the Sharpe. So are you even concerned about the Sharpe so to speak. I guess we’ve seen many of our peers have become somewhat, and many of them actually, maybe to my surprise a little bit, have been very open about it in our conversations, where it is kind of a business decision that they want to put something in that makes it easier to hold and so on and so forth. We translate that in a simple way just saying, yeah, okay, we do pay attention to the Sharpe. So my question to you is, is that something that even plays a role from your perspective?

[00:13:23] Matt

Well, for sure it does. We definitely want to have a reasonable Sharpe. We don’t want to produce something that’s extremely defensive with a negative Sharpe. That’s not going to be attractive to anyone, or even close to zero Sharpe. Theoretically, that would still be of benefit to many people’s portfolio but most people are not going to go for that. So we do want to produce a reasonable Sharpe. Part of the way that we do that is by having a very broad universe. But at the same time, we don’t want to sacrifice the defensiveness in order to maximize the Sharpe.

So like I said earlier, there’s a strong temptation to add strategies to the mix that are value oriented, things that are almost mechanically negatively correlated with trend following, and that’s going to bring up your Sharpe ratio but at the expense of the defensiveness and that’s something that we don’t do. We have the luxury, I guess, of being a relatively small part of a broader firm, most of it is not quant investing.

So we can wait through periods, fallow periods in the trend following space when it’s not really doing that much and still offer that to clients and try to help them understand that this is a long-term allocation and other parts your portfolio should be focused on absolute return, but this part of the portfolio, the real valuable part, is the defensiveness and the diversification.

[00:14:46] Niels

Yeah, I think that’s a really good point, actually, the fact that your Sharpe also is very different compared to the other managers that we tend to come across. Now, one of the things that I’m hoping you can help me out with here, Matt, is that when I do my travels and especially at the moment where people are actually interested in talking to trend followers, which is nice. One of the things I really see a lot of people do when we have these conversations is they very much look at trend following as a trade. They even go as far as thinking that if it did well in 2022, it’s not going to do well in 2023, and so on and so forth.

Can you help us with some of your thoughts and narratives around how can we get investors not to look at trend like that? How I tend to go is that this part of their portfolio, they really need to look at in a different light. But I don’t want to give my narrative. I’d much rather hear your narrative in terms of how do we help people overcome that natural bias towards how they look at trend following?

[00:15:06] Matt

I think the answer is that there’s no real good solution to that. Part of the reason that the returns are persistently so high, in something that’s so valuable in your portfolio, is exactly the reason you’re stating. People just cannot stomach it. It’s doing something generally very different from how most people think about investing. Most people just will not be able to hold it long term no matter how much education they get. But I think you have to try and you have to explain that.

It’s very hard to time trend following. You need to think about it as this unusual thing that you’re allocating to that over time will produce returns and be diversifying. But just simply, I don’t think everyone’s going to get that.

[00:16:42] Niels

Yeah. Absolutely.

[00:16:43] Alan

If I could jump in on that, I think you’re right. We all sense that in our interactions with investors. Which bit of it do you think is it that’s hard to stomach? Is it just prolonged drawdowns, or a seemingly random return profile, or the fact that you can have a drawdown when equities are doing well and there’s a perceived opportunity cost of holding it? Which aspect of the return profile do you think is difficult to endure?

[00:17:16] Matt

I think it’s all of those. I think the basic premise of trend following is the opposite of what most people think about for investing; buy high, sell low type of thing. Most people think it makes no sense. Something that’s an unusual diversifying strategy, with low correlations with almost any other strategy out there, by definition, that means it’s going to have drawdowns at weird times and people are going to be like, “What the heck is going on with this strategy? It makes no sense.”

But they don’t really think that if they’re expecting something that’s a diversifying unusual strategy, it’s naturally going to have some drawdowns and those drawdowns are probably going to be at weird times. You should embrace that. But it’s just very hard to do. People are more comfortable with things that have drawdowns at the same time as the other stuff in their portfolio because it kind of makes sense, even though logically speaking, that’s not what you should want.

[00:18:08] Alan

Yeah, that makes sense. It doesn’t solve the behavioral challenge. Just going back to portfolio construction and how you’ve built the program, I guess if you want to build a trend program that has that stronger defensive characteristic, there are certain things you can do, at least when we’ve spoken to other managers. So one might be trading speed, maybe want to be a bit faster, particularly in equities at turning points, or else possibly, some managers might say don’t trade equities at all if it’s going to be used alongside equities, or else capping the equity beta.

But just curious, what are the levers that you’ve built in terms of developing the program to give it a bit more defensive-like characteristics relative to other managers would you say?

[00:18:59] Matt

Well, the first one I’ve already mentioned, at length almost, is that you don’t want to reduce that by adding diversifying strategies to the portfolio. Then there are a couple of things that we do to actually enhance that behavior. The first, which you mentioned, is that we’re relatively fast. We’re not the fastest out there, but relatively fast. One thing that we’ve convinced ourselves of, looking back over decades of backtest data, is that doesn’t really change your absolute expected return.

I think some years are good for slow trend followers, particularly last year, some years are good for fast trend followers, but on average, I don’t think you should expect a real return difference. But what we do see very consistently is that if you’re a fast trend follower, you’re more defensive so the returns conditional on being in some sort of market drawdown tend to be consistently higher, which I think is logical to expect. If you go back to the option analogy, if you’re a fast trend follower, you’re replicating shorter maturity options, which by definition have higher convexity, higher gamma, higher payoffs in tail scenarios.

So, that’s something you should expect. The trade-off is that you have higher turnover, more false starts and you need to be very careful about execution costs there. But if you’re able to do that, I think that consistently makes it more defensive. The other thing that we do is that we have an asymmetric approach to the equity risk in our portfolio. By equity risk, I don’t mean just outright equities but anything that we view as correlated with equities.

So long oil, long base metals often look like equity risk, long emerging market currencies, all of that stuff, we limit our potential exposure to these risk-on assets, which just naturally makes you more defensive and then also mitigates the potential problem you have like the February 2018 example that I always like to talk about where sentiment reverses overnight. That was seen when we had those VIX ETN blow ups and we went into a bear market very quickly.

In that sort of situation, coming off of a risk on rally, trend followers can be very much on the wrong side of that and it can be very embarrassing when that reverses. So limiting your equity beta (positive equity beta) exposure, can mitigate that to some extent.

[00:21:16] Alan

You said you’re relatively fast, not very fast. Can you give us a sense of in terms of average hold period, how fast? How would you describe how fast you are?

[00:21:28] Matt

Our average holding period is, say, one to two months. So it’s, I think, considerably faster than the average of which is maybe closer to six months, but we’re not talking about intraday trend following for the most part.

[00:21:39] Alan

Sure. I can understand how being faster will get you short when equities turn down. Is there a risk when you’re being faster, in the midst of a prolonged equity bear market, that you might get shaken out of that trend? We’ve had a couple of bear market rallies in equity that counter trend moves. How do you protect against that given that you want to maintain that overall defensive characteristic?

[00:22:09] Matt

Well, I think that just comes with the territory. I don’t think there’s any easy way to solve that. Indeed, we did see that a couple of times last year, particularly in the summer, when we had that bear market rally that quickly reversed. So we got out of our short position there, which was a bad thing to do, but there’s no way to really see that in advance.

[00:22:28] Alan

Okay. Niels, anything on this? I have a follow up question, but it’s more on multi-asset.

[00:22:34] Niels

Yeah, absolutely. I thought it was fascinating, this thing about limiting the risk-on side of things, because from my experience, I’m not a quant, but from my experience, I sit back with this understanding that when you look at trend following and you look at trades, long trades have always historically been a lot more profitable than short trades, short-sided trades.

So is that something you take account of to some extent, because on one hand, you want to limit those risk-on types of trades, but at the same time, you want to probably let your profits run because they can become quite profitable at times. So how do you balance these two slightly opposing forces?

[00:23:23] Matt

Well, I guess we balance it by having a very broad universe. I think once you have a very broad universe, you can find things that aren’t super correlated with equities, generally speaking, that have very large trends that make up for potentially what you’re losing in being more short biased in other assets.

For instance, one of the biggest trends I think we’ve seen the last few years is the European energy crisis, which is obviously a tragedy for many people. But on the trend following front, that was a very large opportunity and started out before the Ukraine invasion.

That was a very small market that I think many people didn’t really pay attention to and certainly not really correlated with equities. So that trend is something that really drove a lot of P&L, and that’s just an example of things that you can find globally. So yeah, it is true that if you’re just biased short equities, that’s probably going to be a negative proposition. But if you have this more broad approach, I don’t think it costs you much.

One thing that we’ve found in our research, which is kind of an interesting observation I think, is that if you have a plot of ex ante equity beta the trend follower. So, how is it positioned right now, negative equity beta or positive equity beta? That’s the x axis, and then on the y axis, you’ve got subsequent realized return. In contrast to most other assets out there, there’s no correlation between the two. It’s just like a random scatterplot.

That’s pretty unusual because whether you’re talking about high yield bonds or loans or real estate or equities or what have you, if you have a higher equity beta, you expect higher returns, generally speaking, because the equity risk premium drives a lot of returns out there. That’s exactly what you’re talking about. Trend following is unusual in that there’s no correlation between the two. So really, it means that when the trend follower is naturally defensively oriented, you really haven’t historically expected lower returns going forward.

Conversely, when you’re risk-on, you haven’t expected higher returns going forward. So I think that’s what drives this and that’s what also drives our belief in the equity beta cap, because the way that beta cap effectively works is we have lower risk in the portfolio when it’s naturally risk-on and we have higher risk in the portfolio when it’s naturally defensive. That’s exactly what you would want in the asset allocation context if you were doing that very dynamically.

[00:25:47] Niels

Makes sense.

Another thing I wanted to ask you in that relation, when you want to have the profile that you’re seeking, do you have to generally look at correlations and volatility differently? Do you think that maybe trend followers who have more (let’s call it) traditional type focus in terms of absolute returns? It’s horrible question to get because we don’t know, necessarily, what other people are doing. But when you think about how you think about volatility and correlations, do you have a different approach to it?

[00:26:23] Matt

No, I wouldn’t really say so. You need to measure correlations in a robust way because if you’re having an equity beta cap, that means you have some view of the equity correlation in these different assets. In order to make that a robust concept, you need to have a robust measure of correlation. That’s not always easy. You need to have a fairly long lookback window to measure that stuff in a robust way in contrast to measuring volatility of assets, which require fewer data points just statistically speaking.

So that means that you can’t be super dynamic in changing your assumption about, for instance, the correlation between stocks and bonds. You need to have a fairly slow-moving assessment of that. But generally speaking, I don’t think our measurement of volatilities and correlations are particularly unique.

[00:27:08] Niels

Yeah. I know this is one of Alan’s questions but it’s in the same section so, because he allowed me in here, Matt, I’m going to ask it and that’s machine learning. Is that something you entertain?

[00:27:23] Matt

We’ve looked at it a little bit in other contexts. For instance, we have equity strategies that do some things based on machine learning, but we haven’t found any obvious uses in the trend following space. I think it’s more of a buzz topic than something that’s necessarily going to pay off in real life. That’s my view. But I potentially am too negative on it.

[00:27:45] Niels

I actually think you probably sum it up pretty well for most of the people we’ve talked about. Even those who say, yeah, we use it a little bit, they all start by saying we use it very cautiously and we always see it, so it’s not a real let it loose kind of thing. So anyways.

[00:28:06] Alan

Just following up on this topic around building this very defensive orientated trend program. I’m just curious about your thoughts on a couple of other asset classes. You obviously talked about limiting the exposure, not just equities, but equity-like products as well. What about the role of bonds in the portfolio? You might have thought, for many years, if you wanted to have that defensive property, you would just lever up on long bonds at the first sign of equity downturns.

Would that have been a part of the approach? The fact that we transitioned into a prolonged bond bear market last year, did that create any kind of challenge in the portfolio or do you keep bonds neutral in that calculation of the overall equity beta?

[00:28:57] Matt

Well, first of all, I should say that we don’t really do anything discretionarily. So we don’t take a view that things have changed in bonds, now the correlation is flipped, we’re going to treat it differently. We allow the models to measure that, in real time, and evolve as appropriate. But yeah, bonds definitely do factor into our equity beta calculation. As you mentioned, for many decades, they looked very defensively oriented, and being long bonds would look like negative equity beta to us.

So to some extent, that would allow us to be, if we were long bonds, that would allow the portfolio to be longer equities as a result and still be below the equity beta cap. That is maybe starting to change. But as I mentioned before, it takes a long time to measure correlations robustly, so I think that has still not changed in our assessment of correlations. We think they probably still have the same behavior roughly speaking that they’ve always had in crisis events, and that’s factored into the equity beta cap. But we don’t do anything. We don’t make any discretionary calls.

One thing that I do think is interesting on this topic which is kind of an aside is we heard many criticisms over the years before the last couple of years that trend followers just profited from being long bonds and is there anything else to it? So I think last year and a half or so has been a vindication on that front that yes, it does do something different. In fact, it can be very diversifying to bond portfolios in many cases. I think that’s very interesting, very good for the space to have seen that episode.

[00:30:35] Alan

The second area I wanted to get your sense on, and again, it was something that probably didn’t work as well last year. But I’ve heard of strategies like this building in a long VIX, dynamic long VIX, or a long vol strategy as part of the protection mix. Is that something you’ve looked at? Again, obviously, that type of thing didn’t work as well last year. So I’m curious to get your observations on that as part of the defensive mix.

[00:31:02] Matt

Well, I guess that gets back to our philosophy on pure trend following. We do run exactly that type of strategy, a long VIX strategy, in our other portfolios and our multi strategy quant portfolios that are not supposed to be pure trend. We do see value in that for exactly the reasons you mentioned. That can be very hedging to more risk on type strategies, short vol in particular, but we don’t think that it necessarily belongs in this type of portfolio. In fact, it would not be super diversifying to the rest of the trend following portfolio. So in that sense also, it doesn’t really make sense. So we really don’t include things like that here.

[00:31:39] Alan

Maybe taking a step back to look at the role of all of these in a larger portfolio, you’ve got equity risk, which we’re not really talking about. But then if you’re thinking about diversifiers and defensive strategies and risk mitigators and protection, historically, it has been long duration trend following or tail risk. If you were constructing a portfolio of equities plus some mix of those, how would you think about what’s the right mix?

[00:32:08] Matt

That’s a very tough question. I think you would want a mix of a number of things, because all the things that you’ve mentioned either cost money, like tail risk hedging, which naturally limits how much you can have there, or doesn’t give you guaranteed protection. Like duration last year obviously didn’t give you a prediction. Trend following happened to work last year but it’s certainly not guaranteed either.

Other things that people have talked about are like long Japanese yen. That often is kind of a flight to quality type asset. That also would have been terrible last year. So I think you really need a mix of all of these different things if that’s your goal to produce protection. But generally speaking, for an absolute return mandate, I personally would want to limit my exposure to these tail risk hedges that cost money, because really, your diversification should be over time and through asset allocation rather than paying actual money to hedge the downside.

That’s why I think trend following is very attractive because it pays you money to do that.

[00:33:09] Alan

Okay. Maybe just moving on to research more generally, can you give us a sense, obviously, you’ve been involved in the strategy for I think you mentioned 15 years or so. Has the strategy evolved a lot over that time period? What have been the big changes that you’ve initiated on the back of research?

[00:33:33] Matt

I guess it’s both. The strategy has evolved a lot in some ways and a little in other ways. So in terms of the pureness and the basic strategy design, it’s evolved very gradually in relatively small ways over time. It’s always had a focus on being fast and defensive, and that we’ve preserved throughout the whole history. In ways where it’s changed a lot is the evolution of our universe. I mentioned this earlier, but we think that having a broad universe is extremely important and that’s something we focused a lot of our efforts on.

So one thing that is very unusual about trend following, shocking almost in contrast to many other strategies out there, is that most of your trades are losers. The big, sustained profitable trends are rare. Only about a third of the trends that we identify, turn out to be profitable but they make a lot more than what you lose on the losers. That’s what drives the long-term payoff of the strategy. But what that means is that diversification is great in every investment portfolio, but here, it’s absolutely essential because you need to cast a very wide net in order to capture as many of those rare trends as possible.

I already mentioned the European energy crisis, which I think for most US investors, at least, was something we never even thought about before the Ukraine crisis. There are other examples than that. Hungarian interest rates starting in the middle of 2021 far before developed market rates really established a trend upward were ticking very steadily upward and that was a very profitable trend.

There are examples like that where there are big trends in strange markets that a lot of people don’t really think about. You really need to have exposure to those things in order to make money over time in a consistent way. So that’s been a big focus of ours.

[00:35:29] Niels

Let me stay on this a little bit. I wanted to ask you, Matt, in terms of people often ask, how are you different to other managers? As I said before, it’s not that we really know, detailed, what other managers are doing, but we have to give some kind of answer. I have a couple of observations I’d love to hear your thoughts about. One is, in terms of the trend following we all do, and let’s just say there are five or six different. You have your moving average type strategies, you have your breakout price, volatility, time series momentum, whatever.

From your research, do you think that actually the type of trend following makes a big difference which one you choose or which mix you choose?

[00:36:14] Matt

This is a question I get all the time. I think I have a view that’s maybe an outlier of some sort. I really think that makes very, very little difference. The technicalities of your rules, breakouts or moving average crossovers, that’s really just in the weeds and makes no real difference. The big difference on this topic is the speed of the ultimate signal. So you could do a breakout that’s very fast. Compare that to a moving average crossover that’s kind of the same speed, those are going to give you very similar performance over time. So I try not to get bogged down into the micro details of the exact implementation.

[00:36:50] Niels

Okay. I’m glad you said that because that’s exactly the same type of answer I’ve been giving for the last 30 plus years. Now, the next question would be, I completely agree also on trend speed, for sure. This is a discussion we haven’t had on the podcast on and off for a few years. I get the point about diversification and I get the point that some managers were early to move into trade hundreds of markets instead of just the classical 60, 70 markets.

However, I’ve also been on record saying when I look at the numbers, if you zoom out, I really don’t see a big difference. Sure, year by year, there can be a year where some types of portfolios, let’s call them that, do better. So I wonder whether we agree on this as well that there may not, in the bigger picture, be such a big difference whether you trade 300 markets or whether you trade 70 markets as such, except that it can be different from year to year, of course. Is that a fair statement or do you have an argument to the contrary?

[00:38:00] Matt

I think you need to look at it in a more isolated way. So instead of just comparing funds where they say they trade 300 markets and potentially they have very little exposure to some of these weird markets, you need to just focus on the performance of the exotic markets compared with the performance of the more vanilla common markets over time. I think if you do that more apples to apples comparison, what you find is that the performance of the more exotic markets has been more consistent, historically, and has suffered less degradation of the Sharpe ratio, over time, than the more commonly traded markets.

I think the more commonly traded markets, they do have more people doing trend following on them, they’re more liquid, more arbed out, and I think the trend following performance has definitely degraded there. We saw astronomical returns in the ’90s and early 2000s, and that has, I think, generally declined, but you haven’t seen nearly as big a decline in the more exotic, hard to access markets.

[00:39:04] Niels

Okay, fair point. Another thing that we like to discuss from time to time is this. It’s not new but it came, certainly, to the fore last year again, and that’s trend replication. I have no idea how your products are structured, but when we spoke with Yao at AQR, I did say to him that they were very early on in bringing out low fee, flat fee type products for trend. What it’s led to, I don’t want to say that they caused it or that they said it, but I seem to remember at least that it led to maybe the impression from investors that trend following is a pretty easy strategy, so you don’t have to pay much for it.

So I wanted to ask you a little bit about that in a sense, whether it is easy, so to speak, and therefore it shouldn’t cost much. But also trend replication, which is then the next iteration of this is people who just look at linear regression of managers’ performance to see if they can create a portfolio that mimics that also became quite popular last year. Since I’m not a quant, I don’t know what risks might lie in these types of models. I’m curious to know what you think about that kind of replication of a trend following strategy?

[00:40:27] Matt

On the first point about it seeming simple, I have had this question a number of times and I always feel vaguely offended by it. Because it certainly does sound very simple in principle, but the devil is in the details. Especially if you’re trading a large number of markets, especially exotic markets, where the data is difficult to get properly or get at all, there really is a lot of work involved in running this in a systematic way.

You can’t just have a spreadsheet and say, “Oh, oil seems to be going up, let’s buy oil.” It needs to be done much more professionally than that. On the trading front, that needs to be done very carefully so that you’re not losing all your alpha in transaction costs and fees. So I really think it’s not something that can be easily replicated, especially if you have a broad universe. On the second point, in terms of replicating it, yeah, doing linear regressions on fund returns, I guess you could capture the positions that funds had over the sample of the regression, but these positions are fairly dynamic, especially the way we do it.

So I don’t think that’s really going to work very well. You’re certainly not going to catch these turning points. I haven’t really tried to do this myself or looked at the data, but I would suspect that the big turning points like in the COVID crisis in March, you’re not going to have the kind of defensive behavior that you hoped to have because you’re going to be very slow. Not to mention the fact that your regression is going to miss unusual things like Hungarian interest rates or European energy or something like that. So I don’t think that’s really a smart way to go about it.

[00:42:10] Niels

Sure. I appreciate that. Alan, where do you want to go next?

[00:42:13] Alan

I just want to pick up a couple of points on trading speed again. I think what you’ve said is your performance for trend following is consistent if you’re fast or slow. From the other managers we’ve spoken to here and some research I’ve seen, there was a sense that fast trend degraded somewhat over the years, that maybe it was more effective back in the 1980s and ’90s and more difficult now.

Would you agree with that? What allows you to be still effective at it? Is it really good execution or is that degradation, is that for very fast trend would you say? What’s your perspective on all of that?

[00:42:54] Matt

I definitely know what you’re talking about. One of the benefits of this type of strategy is that you can backtest it, back a long way. We can easily run decent backtests back to 1960 and I’ve definitely seen what you’re talking about where there were periods of time, particularly in those earlier years where fast trend looked better and then in the recent 20 years, which a lot of people look at, slow trend looked better.

But I think if you look at it carefully, it doesn’t look like a very robust conclusion that there was some kind of level shift down in the returns. It’s more random noise based on what has happened. We’ve had a strong bull market in equities over a good chunk of the last decade or so. That’s been very strong. That doesn’t mean that that type of pattern is going to persist going forward. In particular, interest rates globally steadily declined over the last two decades or so. That’s something that would bias you toward slow trend following. But I don’t see any robustness in that data to say that something has dramatically changed.

In fact, I would expect the opposite from a theoretical perspective. If more money went into the space, you’d probably see a degradation in the slower approach and it would be pushed faster and faster. So it just doesn’t agree. My conclusions from the data is that it’s not robust and it doesn’t agree with my theoretical view on it either. So I don’t generally believe in that. Without casting aspersions on other people, I think it’s very convenient to assume that if you’re a big trend follower, it’s better to be slow because that’s also much easier to do.

[00:44:31] Alan

You can trade more assets, I guess, as well.

Secondly, you emphasized, I suppose, what people would call alternative markets, European energy, Hungarian interest rates. I’ve heard from other managers that when they applied their systems to those markets, they have to trade slower because (maybe they don’t have to) the trading costs are higher. So if you’re trading fast in those markets, it could be prohibitive. Is that your experience or do you apply the same models across all markets?

[00:45:01] Matt

Generally speaking, we tend to be very uniform in our model application because we want to be very careful about data mining. We don’t try to customize our models for the difference between corn and wheat, for instance. We think that it’s hard to come up with robust conclusions on that and it’s very subject to data mining, especially because this is a pretty non-intuitive strategy.

But like you mentioned, we do recognize some things that are guaranteed, and that is that trading less liquid instruments fast is going to cost you more than trading them slow, and that’s something that is not random, that’s guaranteed. So, to the extent that we do vary the model application across the universe, it’s to do exactly what you suggested. We tend to be slower on the less liquid markets and faster on the more liquid markets with the overall balance that the portfolio as a whole is relatively fast.

[00:45:53] Alan

Okay. Again, on the topic of speed, it’s quite interesting. You occupy an interesting space in that a lot of the trend followers are medium term trend followers, and then within managed futures, you have what you call short-term traders. They use a mix of short-term momentum and mean reversion. But I guess they could be trading fast breakout, which might be intraday to a few days.

They do emphasize, I think it’s fair to say, their historical performance has often done well in rising volatility periods and would have that defensive property. I’m just curious, have you guys looked at that very fast space as much? What’s your thinking about having more allocation to even faster systems or fast momentum systems?

[00:46:45] Matt

Yeah, we definitely do look at that. We have allocations to some of that in our multi-asset strategies that are not pure trend following. We do think there’s value there and opportunities, it’s just not very scalable. So, if you want to make a product that’s reasonably scalable and provides a lot of benefit to a lot of investors, that’s just not going to be a huge contributor to it.

[00:47:06] Alan

Okay. So you’re very much trying to construct a portfolio that could run $5 or $10 billion I guess, something like that, is it?

[00:47:14] Matt

Well, yeah, we have $5 billion already. We’re not trying to be the biggest trend follower out there. I don’t want to say that because we also think there’s value in having a diverse universe and that’s going to put a limit on how big you can get. But one of the benefits of trend following, in terms of leverage, is that you’re taking directional bets on things. You’re not doing relative value between different interest rate markets or something like that, which naturally requires higher notionals to get to a certain risk target, and then as a result of that, has a smaller capacity limit. Trend following is very capacity efficient because you’re just taking outright bets. So we think it’s relatively large capacity. But like I said, it’s not unlimited because we want to preserve a decent amount of exposure to these unique markets which really add a lot of value over time.

[00:48:04] Niels

I want to ask a couple of questions that actually got sent from a large allocator in the space. So I just wanted to hear your thoughts about it. Some of the stuff, we’ve touched on before, but nevertheless, I promised to bring it up with an expert like you. So, he writes to me, “If the ultimate objective of investing in trend following is to improve the compounding of your multi-asset portfolio while having tolerable drawdowns, A, should you cap the trend follower’s ability to go long equities if it allows you to maintain both a larger equity exposure outside of trend as well as a higher allocation to trend following?” That was his first question.

[00:48:48] Matt

Well, we do like to cap equities. In terms of capping them at zero and allowing no equity beta in exposure, that starts to be very constraining in the model and does reduce returns pretty significantly. We do that for separate accounts for people who want to very defensive profile, but in general, for general use, I don’t think that really is the optimal approach. So I think what we’ve landed on is a decent constraint on equities, but not too tight that it really reduces the return potential.

[00:49:18] Niels

Okay, that’s fair. Then he goes on to ask, maybe inspired from the AQR episode, I’m not sure. “Should you introduce non price-based trend signals i.e. fundamental trend indicators if they exhibit more consistent crisis alpha?”

[00:49:36] Matt

I certainly am a fan of adding non price-based signals, but only if the key point is if they exhibit that crisis alpha. So we did spend a decent amount of time investigating just this, price momentum on like macro data releases, CPI, and GDP, and all those sorts of things across 25 different countries and we found very good consistent returns over time.

But unfortunately, probably because economies evolve relatively slowly, that model acted like an extremely slow trend follower with almost no defensiveness in the return so it was very good in absolute returns but adding it to the portfolio would have reduced our defensiveness. So we run that in other portfolios but not in our trend following portfolio. I’m a fan of doing that if you can find it, but I wouldn’t add it if it’s going to degrade your defensiveness.

[00:50:32] Niels

Sure. Okay. No, that’s fair. Then he has a question saying (and I’d love your thoughts about this), “How many trend following strategies do you really need in your portfolio?”

[00:50:42] Matt

Well, that’s a tough one as well. I think the answer is less than a large number. Because one thing that’s kind of interesting about statistics is that if you add these positively skewed strategies doing slightly different things, if you add enough of them, because of the central limit theorem, you’re going to get something that’s more normally distributed.

So you don’t want an infinite number of these things doing an infinite variety of parameterizations because that’s going to defeat the purpose, but probably, you want some diversification of styles to give you a more balanced portfolio. So I think the number is probably a small set of different strategies.

[00:51:18] Niels

Yeah, it makes perfect sense. Final question that he wrote was, “If trend following is your first port of call for diversifying a traditional 60/40 portfolio, what’s the second port of call?”

[00:51:30] Matt

That’s another tough one that gets back to the other question. Historically, duration has been very good, stuff like long Japanese yen, and maybe at the values right now, that’s attractive, people have used gold. I think those are all much less reliable than trend following. Duration is probably the next best one, but as we’ve seen…

[00:51:50] Niels

What about long vol and stuff like that? Is that consistently enough? I think a lot of people thought, after 2018, that yeah, we definitely need some long vol in our portfolio. But then maybe after last year, they realized that maybe it doesn’t exactly do what you expect it to do.

[00:52:07] Matt

I think it’s expensive to make sense in a general context. I would never pay for that myself.

[00:52:14] Niels

Okay, that’s cool. Alan, where do we go next? We’ve got another few minutes.

[00:52:20] Alan

I had a question around expected returns, but maybe a better question would be, how do you frame expected outcomes for your investors in terms of saying you have a trend program but with a defensive characteristic or a higher defensive quality? Maybe also, from your own perspective, when you were designing it, a lot of people would approach this problem from the perspective of maybe maximizing the Sharpe or the total return. What are you trying to maximize or what metric are you looking at in terms of measuring the effectiveness in delivering the defensive quality?

[00:53:00] Matt

Probably the performance. The Sharpe ratio conditional on being in equity market drawdowns is probably the key thing that we’re trying to maximize. So, not the absolute Sharpe but what’s the Sharpe conditional on being in those environments. But that said, we still want to increase the absolute Sharpe ratio as much as possible and we do things to do that like having a broad universe.

[00:53:24] Alan

In terms of how you describe that to investors, are you saying that the program is an absolute return that should have cash plus x percent expected return and with the additional characteristic of an expectation that it should be positive? Is it significant equity bear markets? How would you describe that?

[00:53:45] Matt

Yeah. Any kind of very significant equity market drawdown, we frequently look at quintiles of equity quarterly returns, and we want to see good performance in that worst quintile (to give you an example of how extreme a situation we’re talking about). So small drawdowns won’t necessarily register but the big ones, the worst ones, we should expect to be positive on average.

[00:54:07] Alan

Okay. Maybe just a question around liquidity, which is probably quite relevant being even a faster trend follower than average. There’s been plenty of chat in the media on liquidity, maybe it isn’t as good as it was, a lot more high frequency traders, etc., and we’ve had a couple of incidents where we’ve had exaggerated moves in markets; gilt markets recently, treasuries back in 2020.

Obviously, you’re trading fast, relatively fast in a market and so liquidity presumably is important. What’s your sense on liquidity conditions? Any notable shifts there over time with a change in the market structure or not?

[00:54:48] Matt

Well, we’re focused entirely on liquid derivatives. So I think the changes that we’ve seen in the markets have been applied more to cash securities where people need a balance sheet to intermediate risk and less so to derivatives trade. So I don’t think we’ve seen a major change in liquidity in the assets that we trade.

[00:55:10] Niels

You can of course interrupt me, Alan. But usually at this stage, Matt, I’m allowed to ask a few more fun and general rapid-fire questions. So the first one today would, of course, be in light of the fact that there is someone who used to work at your firm that wrote a book called The Bond King, I was wondering whether you’re publishing a book called The Trend King.

[00:55:34] Matt

No. Bill Gross is a towering figure. I would never put myself in that category.

[00:55:42] Niels

Fair enough. One thing we actually ask all of our guests, and it’s a little bit of a trick question, but it’s not really a trick question. But it is, what’s the one thing you hear about trend following that you disagree with the most? So things that you maybe hear people say about trend and where you say no, I definitely don’t agree with that. Is there something that springs to mind?

[00:56:05] Matt

Yeah, something that definitely springs to mind is the headlines that I see periodically about trend followers being long equities or short equities, and potentially, if there’s a change in the horizon, it’s going to affect the S&P 500 or 10-year bond futures. I think that idea is way off. I think trend followers are far too small to affect those very, very liquid markets. So I think it’s more of clickbait than anything else.

[00:56:32] Niels

You know what? That’s the first time someone had brought that up, but I couldn’t agree more. I’m not saying that. That’s the one I would pick, but it’s definitely in my top three. Of course, you don’t know this, Matt, but on this day where we are recording, Alan and I have just recorded our weekly conversation and we spend a whole 15, 20 minutes discussing the latest Bloomberg article that goes exactly to that point, but also some analysis from another firm about the footprint of CTAs. So anyways, I love your answer and I’m a fan of that choice for sure.

Finally, and then if there’s anything else you want to bring up, Matt or Alan. Now that we are going into 2023, clearly 2022 was a year where people were reminded of the attributes that we bring to a portfolio. What are you most excited about when you look into the year? Do you have any concerns in terms of how the financial world is behaving?

[00:57:38] Matt

It’s always hard to predict what trend volume is going to do. I would definitely never predict that. But one thing that I am kind of heartened by is the reduced presence in global markets of central banks. There was a very fallow period in trend following, ’15, ’16, ’17, ’18, where trends that tried to get started were kind of stopped in their tracks by very active central banks globally coordinated in some sense, and that is really changing.

So with the inflation that we’ve seen globally coming out of COVID, central banks now have their hands full with multiple potential problems to deal with and they really can’t be as effective or as active as they have been historically. That creates a riper environment for trend following.

Even the BOJ, we’re seeing signs of life there were potentially their extraordinary measures might be removed in the coming year, and that’s really a very interesting thing to me because that’s been in place for a very long time and that market really has not done much over that period of time. So to the extent that themes are less controlled and free to flow to move, that allows more trends to form, and to that extent, I think it’s potentially a very good environment.

[00:58:52] Niels

Yeah, I agree. That’s a great point.

Matt, is there something that you feel we forgot to ask you or is there a particular topic that you, internally, or with your clients and prospects like to talk about that we didn’t have in our lineup today, something you want to mention before we start to wrap up?

[00:59:12] Matt

No, I think we’ve pretty much covered the main things. The most important thing I would mention, aside from the defensive profile of trend following, is just the universe. It’s one thing I think people should really focus on, making sure when they’re investing in trend following that they’re getting material exposure to a very broad universe, because as I mentioned, trends being relatively rare, your best chance of success is having exposure to these strange corners of the market.

I don’t mean just having them in the portfolio in some token amount but having material exposure to things like European power or Hungarian interest rates when trends pop up there. So I think that’s a very important thing to keep in mind.

[00:59:51] Niels

Yeah, that’s a great point. Alan, anything you want to…

[00:59:54] Alan

Maybe just one final one, which may be a little bit left field. Obviously, you mentioned The Bond King. If you mentioned the word PIMCO, people obviously think in terms of a bond shop. I’m just curious, what’s the experience of running trend following like in a bond shop from the perspective of is that more challenging when you go to meet clients?

Are people surprised to hear you’ve got a big managed futures business? Are there benefits to that as well? I guess there is the brand that you have. But just curious, any observations around that?

[01:00:34] Matt

I would say, generally speaking, it’s been very additive in a couple of different ways. Generally speaking, investors here are very value oriented so they provide a healthy dose of skepticism about what we’re doing and keep us on our toes, if you will. On the more substantial front, we get all kinds of ideas from the discretionary traders here. I already mentioned, a couple of times, Hungarian interest rates. We trade a very broad set of instruments.

There are many liquid fixed income markets out there, and we have the benefit of being able to add those to our portfolio in a pretty easy way because people have already done the legwork of figuring how to trade those, we’ve got experts to trade those, we’ve got all the data, which is often hard to come by, and that’s just one example. So generally speaking, being part of this broader organization with a very wide footprint has been extremely additive to our approach.

[01:01:33] Niels

Great stuff. I think we’re going to wrap up on that note, Matt. This has truly been fascinating. It’s been delightful. Thank you so much for being on the podcast and for sharing your thoughts and insights with us. Hopefully, we can do this sometime again in the future.

To all of you listening today, I hope that you were able to take something from today’s conversation on to your own investment journey. If you did, please share these episodes with your friends and colleagues.

From Alan and me, thanks so much for listening. We look forward to being back with you on the next episode of Top Traders Unplugged as we continue our deep dive into the CTA industry. In the meantime, go check out the show notes for this episode and all the other resources you can find on our website. Of course, not least, take care of yourself and take care of each other.

[01:02:19] Ending

Thanks for listening to Top Traders Unplugged. If you feel you learned something of value from today’s episode, the best way to stay updated is to go on over to iTunes and subscribe to the show so that you’ll be sure to get all the new episodes as they’re released.

We have some amazing guests lined up for you and to ensure our show continues to grow, please leave us an honest rating and review on iTunes. It only takes a minute and it’s the best way to show us you love the podcast. We’ll see you next time on Top Traders Unplugged.